Two tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. generate_offer has 6 parameters with type and description, but lacks enums for constrained fields like 'urgency'. check_offer_status is minimal (1 param). Neither tool documents output structure, error handling, or recovery guidance. Descriptions are adequate (50-80 chars) but lack WHEN/WHY context. No tool annotations (readOnlyHint present in metadata but not in schema). Parameter descriptions are generic and lack format/constraint details (e.g., 'urgency' has no enum or valid values listed). No pagination, no batch operations, no idempotency guarantees documented.

• Document output schema for both tools. Example for generate_offer: {quote_id: string, total_cost: number, currency: string, estimated_delivery: string, settled: boolean}. This enables the LLM to extract quote_id and pass it to check_offer_status.
• Add enum constraint to 'urgency' parameter: enum: ['standard', 'express', 'overnight', 'custom']. Include in description: 'Speed requirement: standard (5-7 days), express (2-3 days), overnight (next day), or custom.'
• Expand parameter descriptions with format/constraint details. E.g., 'destination_zip: ZIP code (5 digits, US format)' and 'dimensions: Physical dimensions as 'LxWxH cm' per carton, e.g. '60x40x40 cm'.'
• Add error handling guidance to tool descriptions. E.g., 'If origin is unsupported, the response will list available origins. If destination is unreachable, consider alternative ZIP codes.'
• Add tool annotations to schema: mark both tools with readOnlyHint=true (they do not modify state). This signals to agents that these calls are safe to retry.